This was reported by a static analyzer tool run against the Chromium projects, which make use of Mesa. The downstream bug is http://code.google.com/p/chromium/issues/detail?id=84134

The original bug was reported against the Mesa used by Chromium, which is Mesa 7.6, and is as follows:

V501 There are identical sub-expressions to the left and to the right of the '&&' operator: width > 0 && height > 0 && height > 0 mesa teximage.c 2801

void GLAPIENTRY
_mesa_TexSubImage3D(...)
{
  ...
  else if (width > 0 && height > 0 && height > 0) {
  ...
}


However, this bug also exists in HEAD at http://cgit.freedesktop.org/mesa/mesa/tree/src/mesa/main/teximage.c#n2696
in the refactored method:

static void
texsubimage(...)
{
  ...
      else if (width > 0 && height > 0 && height > 0) {
         /* If we have a border, offset=-1 is legal.  Bias by border width. */
  ...
}

The issue is the additional height > 0 check. I'm uncertain if this is a redundant check, and can be eliminated, or if it should be checking depth > 0